For the past 15 years the Biostatistics Shared Resource has played a vital role in the research life of the UWCCC Members of the Shared Resource are an essential component in each of the research program from the laboratory bench through to the clinic. Dr. Mary Lindstrom and Michael Newton specialize in laboratory collaboration. Their contributions have induced a new method for assessing gene loss (Newton) and a new approach to comparing radiation cell survival curves (Lindstrom). Drs. David Demets and Richard Chappell and, before leaving, Barry Storer, play an essential role in the conduct of clinical research in the UWCCC. Stored developed new methods for phase I and II design now being continued by DeMets. DeMets has developed data monitoring methods for phase II and III trials which has been used extensively by researcher at the UWCCC and internationally. Chappell has developed a new model for therapeutic radiation damage to tumors. Drs. Jeffrey Douglas, DeMets, and Polly Newcomb have also been very active in Cancer Control, primarily with population, measurement and analysis issues concerning cancer symptoms such as pain and fatigue and a variety of epidemiological issues. These and other collaborators have resulted in 68 collaborative publications in the four years 1993-1996. One of the strengths of the UWCCC is the involvement of biostatisticians from the initial stages of project development to the completion of the project. The shared resource staff function as collaborative members of research teams drawing a portion of their support from peer reviewed grants. This approach, with its emphasis on collaboration, maximizes the benefit of the resource and helps foster and retain an expert, motivated and innovative staff committed to their specific areas of collaboration. Funds from the CCSSG are critical in maintaining this shared resource, encouraging investigators to seek our support for advice and expertise in the initial stages of a project. The initial stage of a project is critical to get the experimental design established so that the completed experiment can be properly analyzed and sensibly interpreted.